Exploring influencing factors affecting preventive behaviours of Plasmodium knowlesi malaria among at-risk communities: an application of the nominal group technique

Abstract

Introduction

Malaria remains a major public health issue in Southeast Asia, with Plasmodium knowlesi (P. knowlesi) emerging as a growing threat. Despite intensified prevention and control efforts, populations in this region remain highly vulnerable to this zoonotic disease, driven by the complex interactions between primates, mosquitoes, and humans. This study aims to explore understanding of the factors affecting communities vulnerable to P. knowlesi malaria and their preventive behaviours, tailored to the local context.

Methodology

A group of participants with substantial expertise and experience in malaria programme implementation and field operations were assembled. Utilizing the Nominal Group Technique (NGT), a validated and systematic approach for facilitating group discussions aimed at consensus, we identified the essential points to prioritize factors affecting at-risk communities and their malaria prevention behaviours. NGT were conducted through virtual platforms in October 2024, achieving consensus when 80% of the experts concurred on specific ideas. Final rankings were established based on descending acceptance percentages.

Result

Of the 21 items presented for voting, 20 were retained for final ranking, demonstrating a strong panel consensus on key factors influencing malaria risk and prevention behaviours. The voting response rate was 100%, and consensus was achieved in a single NGT round. The 20 selected items, endorsed by the expert panel, were categorized into four overarching themes: environmental risks, human-related risks, policy and organizational factors, and vector-related risks.

Conclusion

These findings provide a valuable foundation for refining malaria prevention strategies. Future research can leverage this consensus to deepen the understanding of malaria-prevention behaviours and enhance P. knowlesi malaria programs tailored to at-risk communities.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12936-026-05792-5.

Introduction

Malaria remains a significant public health concern in Southeast Asia (SEA). Although the majority of these cases are caused by Plasmodium falciparum (P. falciparum) and Plasmodium vivax (P. vivax), the emergence of Plasmodium knowlesi (P. knowlesi), a zoonotic malaria, in the Southeast Asian and Western Pacific regions has caused global concern [1]. Due to variables including population mobility, vector resistance, and the scope of control measures, the incidences of malaria in SEA have varied significantly among countries [2]. In 2023, the World Health Organization estimates as many as 263 million cases of malaria across 83 endemic countries with deaths of more than 597,000 cases [1]. Although cases of P. falciparum and P. vivax malaria are still prevalent in Africa, SEA, the Eastern Mediterranean, and the Western Pacific region of the world, P. knowlesi malaria infection in humans is seen to be increasing and is considered a threat to public health [1]. Globally, there were 3290 reported cases of P. knowlesi infection in 2023, marking an increase of 18.9% from 2768 cases in 2022, with Malaysia continues to account more than 80% of total cases [1]. This rise may reflect improved diagnostic capabilities, ongoing human–macaque–vector interactions due to land-use changes, and the unique zoonotic nature of P. knowlesi transmission [1, 3, 4]. The rising burden and transmission of P. knowlesi present unique challenges for malaria elimination efforts and have implications for malaria free certification.

Efforts to prevent and control P. knowlesi malaria in SEA have intensified and multifaceted, with combination of vector control strategies, improved surveillance systems, and public health interventions to tailored with region’s unique challenges. Many countries, including Cambodia, Myanmar, and Laos, have ramped up insecticide-treated bed nets (ITNs) distribution and promotion of indoor residual spraying (IRS) through national malaria programs, particularly in high-risk rural and forested areas [5]. Efforts to enhance public awareness and preventive behaviour have included educational campaigns emphasizing the importance of personal protection measures, such as using mosquito repellents and wearing long-sleeved clothing [6]. Furthermore, in order to reach people in remote locations and educate them on the disease, many SEA countries have adopted more comprehensive strategies, like community engagement and community health workers [7, 8]. Beyond healthcare and vector control, multi-sectoral collaborations between governments, research institutions, and non-governmental organizations (NGOs) have played a crucial role in addressing malaria [8, 9]. Despite the progress made, populations in SEA continue to face significant vulnerability to P. knowlesi malaria, driven by persistent challenges that must be addressed with greater precision. Those most affected are communities living or working near forested areas, especially individuals in forest-fringe occupations such as farming, plantation work, and indigenous livelihood activities [10]. Unlike P. falciparum and P. vivax, P. knowlesi disproportionately affects adults due to its zoonotic and outdoor exposure patterns [10].

The transmission dynamics of P. knowlesi malaria differ significantly from those of human malaria due to the intricate interplay between primates, mosquitoes, and human. These interactions are influenced by spatial and temporal heterogeneity, making zoonotic malaria transmission highly complex [3, 11]. Mosquitoes carrying infectious Plasmodium species must bite people in order for humans to contract the disease. Being close to infectious vectors is necessary for this, and changes in land use, occupation, and home construction are frequently linked to these changes [12]. Diversity in social context and belief play an important role in malaria exposure especially on perception and practice among rural communities [12]. An additional factor contributing to the disease's incidence is environmental impact. An increased occurrence of P. knowlesi has been linked to the reduction in forest cover, both recent and historical, in the vicinity of rural settlements in parts of SEA [13]. The destruction of natural ecosystems forces mosquito populations closer to human settlements, facilitating the spillover of zoonotic malaria [13]. Given these multifaceted determinants, rural and remote communities in endemic regions face a heightened risk of infection due to their increased exposure to both vectors and reservoirs of the disease [14, 15]. Addressing these challenges requires a comprehensive and context-specific approach that integrates ecological, socioeconomic, and behavioural interventions tailored to the unique needs of affected populations.

The burden of malaria has declined significantly over the past decade, but more work is still needed to close the gaps and eliminate this zoonotic disease. Preventive actions and control have already in place to achieve this target. However, it has been noted that such controls have been implemented in P. knowlesi endemic region but failed to lessen the incidence of this zoonotic malaria [16, 17]. To address gaps regarding P. knowlesi malaria, this study seeks a different paradigm to expand the existing knowledge on this disease. Thus, diverse viewpoints and strategies are required, including integrating insights from local communities and perspectives from experts. To effectively address subjective issues and explore diverse perspectives, the implementation of structured consensus approaches is essential. The nominal group technique (NGT) is one of the primary consensus methodologies that is widely utilized in healthcare research, where structured approach is delivered through a democratic process [18]. The nature of the NGT ensures coproduction of knowledge, mitigates the balanced participation, prioritizing information, and enhances the reliability of collective judgments. Recently, there has been an increased interest in conducting research methods in a virtual environment, where participants engage synchronously through videoconferencing software [19]. We conducted virtual nominal groups with local experts to address our research objective of achieving consensus on factors influencing communities at risk of P. knowlesi malaria infection and their preventive behaviours, adapted to the local context.

Methods

Study site

This study is carried out in Perak, a state of Malaysia on the west coast of the Malay Peninsula. With an equatorial climate, Perak boasts a variety of tropical rainforests. As part of the broader Tenasserim Hills system that links Malaysia, Thailand, and Myanmar, the state's mountain ranges are a part of the Titiwangsa Range and typically experiencing hot, humid, and wet equatorial climate. 1,019,052 hectares make up Perak's forested area, which comprises of forest lands, mangrove and forest plantations [20]. Based on the statistic between 2011 to 2018, Perak recorded as the third highest P. knowlesi malaria cases in Peninsular Malaysia, which make up approximately 18% of total Peninsular cases [21]. Kelantan and Pahang reported the highest numbers and both share land borders with Perak, forming a continuous belt of forested and semi-forested areas [21]. The average age of the identified malaria cases were around 30 to 39 years old, predominantly in males and most had occupations related to agriculture and the forest [21]. The 2024 Perak State Malaria Report shows that Muallim District recorded the highest number of P. knowlesi cases compared to other districts in Perak [22]. In Batang Padang district, P. knowlesi cases increased by 600% in 2018 compared to the previous year, with 27 cases recorded, most of which involved Orang Asli communities with a history of forest-entry activities [23]. A total of 27 patients recorded in that year with predominant Orang Asli communities and had history of jungle trekking. A breeding population of Anopheles leucosphyrus larvae was found in the areas, and nearly half of the macaques that were identified were harbouring P.knowlesi [23].

Study design

This study conducted in October 2024, using the NGT method, which involves four essential stages: silent generation of ideas, round-robin listing of ideas, discussion of the ideas and independent voting by each member (either ranking or rating). These stages are well-documented in the literature, and the NGT process used in this study followed the checklist established by Humphrey-Murto et al. to ensure methodological rigor [18]. NGT was selected because it supports structured and balanced participation and allows clear identification of shared priorities. These strengths are important for understanding community views and behavioural factors in zoonotic malaria [18]. We adapt the NGT process to a virtual environment, allowing participants from different locations to contribute and collaborate online. We summarized our NGT steps in Fig. 1.

Fig. 1.

virtual NGT steps

Expert panels recruitment process

The recruitment of panels for this study followed a structured and strategic approach to ensure that only qualified participants with a level of expertise in malaria research and groundwork were involved. The list of potential panellists was compiled in consultation with the district health offices, with particular focus on Muallim and Batang Padang, which are among the districts with the highest recorded P. knowlesi cases in Perak. The panellists comprised of national and state-level experts as well as individuals residing or working within these two districts, thereby providing both broader programmatic perspectives and insights grounded in local experience. Initially, an email invitation was sent out to potential panels, outlining the key objectives of the study, the various phases of the NGT, and the estimated duration of each phase. The invitation also provided a comprehensive overview of how consensus would be achieved through the NGT process, as well as the contact information of the researchers for any queries or clarifications. Panellists were considered eligible for inclusion in the study if they had substantial expertise and experience in malaria programme implementation and field operation in Perak. Given that prior research has highlighted the susceptibility of NGT studies to subjective biases, we adopted a purposive and snowball sampling approach to select participants. This methodology was implemented to ensure that the sample was specifically tailored to the needs of the study. The selection of panellists was based on their consent for this study, their more than three years of experience working or practicing in health institutions, their administrative position in any health organization, and/or their research on malaria or epidemiology. No formal exclusion criteria were applied; all individuals who met the inclusion criteria were eligible to participate. Traditional NGT groups typically consist of four to seven participants, although larger groups of up to fourteen are also common [18]. For this study, the panel included three public health physicians, three field medical officer, three senior public health assistant, and one entomology officer, all with extensive experience in malaria prevention and control activities in Muallim and Batang Padang district. Ultimately, 10 panels were considered adequate to achieve meaningful consensus and results. The expert panel, therefore, reflected the high calibre and relevance of the participants, ensuring that the study was grounded in informed and authoritative perspectives on malaria research.

NGT process

To ensure clarity and alignment of purpose, we conducted engagement sessions before the main meeting to communicate the study's objectives and goals to the participants. These sessions provided the necessary context for participants to understand the scope and significance of their contributions. Prior to recruitment, all panellists were provided sufficient time and information to make an informed decision regarding their participation, after which verbal informed consent was obtained. This process ensured that participation was entirely voluntary and conducted in accordance with ethical principles. Following consent, a single group session was conducted with all panellists present. This format was selected to optimise engagement by allowing interactive discussion among participants with diverse roles and experiences, and to ensure broad representation across the malaria control service, including both state-level experts and individuals working within the study districts. Prior to the session, panels were encouraged to review the Guidelines for Malaria Vector Control in Malaysia [24] along with relevant literatures, particularly focusing on behaviours or activities that increase the risk of P. knowlesi malaria infection, as well as the challenges in preventing the transmission of malaria [2, 6, 12, 25]. This preparation ensured that the study integrated both evidence and expert opinion systematically.

The expert panels convened virtually using Zoom® which was recorded to maintain a detailed historical record of the discussion and decisions made. This recording serves as a reliable resource for verifying the decisions taken during the meeting. The moderator’s role, undertaken by SAMH, was to ensure that panels stayed on track, remained focused on the topic, and adhered to the structured process of the NGT. At the start of the session, SAMH provided a brief overview of the NGT methodology, offering a clear explanation of the various stages involved. This introduction laid the groundwork for understanding the systematic approach of the process. The panels were then presented with the key question: What factors or elements can influence communities at risk of Plasmodium knowlesi malaria infection and their behaviour in preventing the disease?

The session followed the traditional NGT steps. The first phase involved silent idea generation, where each panel was asked to individually reflect on the question and write down their ideas in isolation. After this time, the idea-sharing session commenced. Each panel took turns presenting their thoughts, with the moderator carefully listing and displaying all ideas for the group to see. Panels were allowed to introduce new ideas during this process but were required to wait their turn before speaking. This step continued until all ideas had been shared, and no more suggestions were forthcoming. Once the list of ideas was completed, the process moved into the idea discussion phase. One item was discussed at a time, ensuring that all participants fully understood each point before moving on. The discussion provided an opportunity for participants to elaborate on their initial ideas, challenge assumptions, and ask questions to refine the understanding of each point. In this phase, similar ideas were grouped together with the agreement of all panel members. This collaborative approach helped streamline the list and ensured that everyone was aligned in their interpretation of the ideas presented. The final step in the NGT process was the voting phase, where each panel privately ranked the ideas generated during the session. To facilitate this, panels were provided with a Google Form® ranking sheet, which allowed them to evaluate each idea on a 5-point Likert scale. The scale ranged from 1 (indicating "not at all important") to 5 (indicating "extremely important"), enabling participants to assign a value that best reflected their assessment of each idea’s relevance and significance.

Data analysis

Once all participants had submitted their votes, the rankings for each idea were compiled. The individual scores were then summed and converted into percentage form to provide a clear, quantitative measure of each idea’s relative importance. This conversion enabled the ideas to be ranked from most to least important based on the collective input of the panel. The final ranked list of ideas was then presented to the group for final discussion and closing session. The goal was to reach a collective agreement on the most important ideas, based on both their rankings and the final group discussion. The summary and calculation of scores were conducted immediately during the NGT session to allow real-time feedback to the panel. All scoring, compilation, and ranking procedures were carried out by the moderator, who was responsible for facilitating the session and ensuring that the analysis followed the established NGT steps.

Result

In just one meeting, we successfully reached a final consensus and addressed the study’s objective, which aimed to identify the factors and elements that influence communities at risk of P. knowlesi malaria infection and their behaviour in preventing the disease. The meeting took place in October 2024 via the Zoom® platform and was attended by all 10 consented panels. The panel comprised public health professionals, entomological officers, and local health workers with in-depth knowledge of malaria ground operations, several of whom also had experience in operational or implementation research. Meeting the inclusion criteria, most panels had over two years of experience and involved in malaria research, with half having dedicated at least five years to malaria prevention and control. Their extensive expertise, particularly in working with remote and rural communities, underscores their deep understanding of the challenges and strategies involved in combating malaria in these vulnerable regions. The session lasted three hours, during which the panels engaged in a structured process to generate and refine ideas for the framework. This efficiency highlights the effectiveness of NGT in producing focused, actionable outcomes within a relatively short period, compared to other methodologies that require prolonged interaction.

During the idea-sharing phase, expert panels initially generated 35 distinct ideas, reflecting diverse perspectives on malaria risk and preventive behaviour tailored to their experience handling malaria cases in this local setting. On average, each panellist contributed between three to five ideas, resulting in a comprehensive pool of items. All ideas were presented verbally, and the moderator recorded them verbatim without discussion. To enhance clarity and organization, these ideas were systematically summarized, categorized, and presented using Microsoft Excel®, ensuring ease of reference and facilitating structured discussions. Each proposed item was then carefully reviewed by the panels, who engaged in detailed discussions to clarify ambiguities, refine concepts, and identify overlapping themes. Several ideas with similar meanings were consolidated into single item, streamlining the list while preserving the essence of each contribution. This iterative refinement process ensured that all perspectives were adequately represented before proceeding to the next stage. Following the review, expert panels categorized the finalized ideas into four overarching themes: environmental-related risks, human-related risks, and policies and organization, with an additional consideration for vector-related risks. This structured classification allowed for a more focused evaluation of each issue within its relevant context. Ultimately, 21 distinct items were selected for the voting phase, ensuring that all key concerns identified during the discussion were retained for further prioritization. Descriptors for each category were developed collaboratively during this stage to ensure that the grouped items accurately reflected the underlying themes prior to the ranking exercise.

For the voting process, each item was assessed based on its perceived importance. The 80 percent threshold was agreed upon at the beginning of the session through a brief group discussion, during which panellists confirmed that this level of agreement would reflect a strong and acceptable consensus. This threshold provided a clear and objective measure for determining the most critical areas requiring intervention. The moderator facilitated this discussion by outlining standard practices in NGT consensus setting, but the final decision on the threshold was made collectively by the panellists. The results of this ranking process are summarized in Table 1, which presents the prioritized items under the four broad thematic categories established during discussions. Notably, out of the 21 items subjected to voting, 20 were ultimately retained for final ranking, reflecting the panel's strong consensus on key factors affecting communities at risk of P. knowlesi malaria and their preventive behaviours. Qualitative discussion during the discussion phase helped refine and strengthen the item list. Most ideas were agreed upon quickly, while minor differences were resolved through brief discussion. Overlapping items were merged, and one item was removed after consensus that it was less relevant. These discussions ensured that the final priorities reflected shared understanding and practical experience. The structured and collaborative nature of the NGT approach not only facilitated the identification of priority areas but also ensured that the final ranked items reflected a well-rounded, expert-driven consensus. By integrating both qualitative discussion and quantitative prioritization, this process offered a comprehensive framework for understanding the complexities of local preventive behaviours.

Table 1.

NGT result

No.	Items generated	Descriptors	Categories	Percentage score value (%)	Acceptance result	Ranking
1	Climate and weather	Climatic conditions such as temperature, rainfall, and humidity, affect the risk of transmission	Environmental Related Risk	47/50: 94%	Accepted	3
2	Season	Seasonal trend such as fruit season affect mobility of community to forest		47/50: 94%	Accepted	3
3	Forest Fragmentation	Fragmented forest increases human-wildlife interaction, such as primates		48/50: 96%	Accepted	2
4	Geography/Landform	Topographical features such as elevation and river system affect P. knowlesi vector distribution		43/50: 86%	Accepted	7
5	Reservoir/Macaque population	Housing areas near reservoir/macaque population increases human-wildlife interaction		46/50: 92%	Accepted	4
6	Ecotourism/Jungle Exploration	Ecotourism increases human-mosquito interactions, disrupts wildlife, and raises malaria risk by exposing tourists to primate habitats		44/50: 88%	Accepted	6
7	Cleanliness in surrounding housing areas	Tall grasses, dirty and water bodies around housing areas favour vector distribution		40/50: 80%	Accepted	10
8	Daily routine	Collecting forest product as daily activities predispose them to P. knowlesi	Human Related Risk	49/50: 98%	Accepted	1
9	Night activities	Hunting or gathering in forested areas after sunset, exposes individuals to these mosquitoes		48/50: 96%	Accepted	2
10	Cultural practices	Common practices like sleeping outdoors or staying in the forest can significantly increase the risk of contracting malaria		48/50: 96%	Accepted	2
11	Community belief	Some communities may not consistently use preventive measures such as insect repellent or bed nets, due to cultural beliefs		44/50: 88%	Accepted	6
12	Health-seeking behaviour	Lack of awareness and not interested in health education/information affecting disease prevention and control in the community		45/50: 90%	Accepted	5
13	Knowledge	Knowledge of P. knowlesi malaria—its transmission, symptoms, and prevention greatly influence individual and community actions		45/50: 90%	Accepted	5
14	Education Level	Higher education increases awareness, health-seeking behaviours, and adoption of preventive measures, reducing the risk of P. knowlesi malaria		44/50: 88%	Accepted	6
15	Community mobility	Mobility especially among indigenous people increases the risk of P. knowlesi malaria by raising exposure to infected areas, facilitating disease spread and affecting healthcare access		42/50: 84%	Accepted	8
16	Hobby	Hobbies like making monkeys as pets may affect vector distribution and increases zoonotic interaction		46/50: 92%	Accepted	4
17	General immunity	General immunity can lower the risk of infection, especially in populations with prior exposure or strong immune responses, but those with weaker or no immunity remain at higher risk		38/50: 76%	Rejected	–
18	Health Access	Health access influencing timely diagnosis, treatment, and the use of preventive measures	Policies and Organization	40/50: 80%	Accepted	10
19	Government Act and Policy	Effective acts, legislation and policy with collaboration with local agencies can reduce the risk of P. knowlesi malaria by regulating preventive measures, funding malaria programs, ensuring healthcare access, promoting public education, and improving surveillance		43/50: 86%	Accepted	7
20	Village head/Community leader’s role	Community leader’s role in reducing the risk of P. knowlesi malaria by advocating resources and coordinating efforts to improve malaria prevention and care		41/50: 82%	Accepted	9
21	Vector Adaptation	Changes in resistance, feeding habits, breeding locations, longevity, or behaviours—can increase the risk of P. knowlesi malaria	Vector Related Risk	41/50: 82%	Accepted	9

Discussion

The virtual NGT demonstrated significant flexibility and adaptability, rendering it suitable for addressing complex research questions. We selected NGT for several key reasons: first, its primary focus on facilitating idea generation aligns with our objective of comprehending malaria exposure and preventive behaviour through the perspective of local healthcare providers; second, its organised, interactive framework ensured equitable and meaningful contributions from all panel members; and third, it fostered constructive debate to reconcile differing viewpoints, addressing knowledge gaps or uncertainties that emerged, thereby advancing systematic knowledge development [18]. The panel members utilised their knowledge and experiences during the meeting, ensuring that insights were preserved, and a substantial amount of data was gathered in a brief period. This study enhances understanding of malaria by drawing on the collective knowledge of panellists representing different roles and levels of experience within the malaria programme and broader health sector. The NGT model systematically documents all ideas, thereby safeguarding against the loss of insights that may arise from the inherent uncertainty of certain panels [18]. Expert panels reached a consensus on the identification and prioritisation of key factors to be investigated in relation to the local community context of malaria. The importance of these factors in influencing malaria-prevention behaviour among communities affected by P. knowlesi malaria studies was developed and integrated into a theoretical framework.

Synthesis of findings

Consistent with prior findings, human and environmental related risks were the most salient factors influencing malaria infection and its prevention behaviour in this local setting of Malaysia. Factors like daily routines and night activities are usually mentioned as individual and interpersonal risk related to P. knowlesi in most studies in SEA, which usually correlated with occupational- and livelihood-related activities [12]. These activities bring them into closer contact with mosquito vectors and macaque reservoirs, predisposing them to such zoonotic infection. Evidence also suggests that communities with limited knowledge about P. knowlesi transmission mechanisms are less likely to implement protective behaviours, which later shaping their daily practices [26]. Cultural beliefs might also influence the perception of malaria, potentially hindering the adoption of preventive measures or prompt treatment-seeking behaviours [27].

Environmental variations including land cover types, climate changes, anthropogenic landscapes, and host distributions have been linked to the geographical distribution and altered transmission patterns of malaria and other vector-borne diseases in SEA [3, 28, 29]. A modelling framework estimates comparable heterogeneity levels of predicted P. knowlesi risk secondary to environmental changes across SEA, with the highest transmission risk of the pathogen is concentrated in Malaysia and Indonesia [3]. Pertaining to the current local setting of this NGT, key environmental covariates affecting human P. knowlesi occurrence may be due to impact of tree loss, distance to the forest, annual precipitation, tree cover, and low elevation (75–345 m above mean sea level) areas along the Titiwangsa mountain range and inland central-northern region of Peninsular Malaysia [4].

Despite efforts by the World Health Organization and national programs to strengthen malaria control, the lack of integrated "One Health" approaches and intersectoral collaboration in creating effective policies remains a significant gap in addressing P. knowlesi malaria [30]. In Malaysia, where P. knowlesi is now the predominant malaria species, aggressive malaria elimination strategies have successfully reduced human-only malaria cases but have left populations vulnerable to zoonotic transmission, which is not addressed by traditional malaria control methods [31]. The integration of P. knowlesi into national malaria policies remains a challenge, as vector control strategies like insecticide-treated nets and indoor residual spraying are less effective due to vector adaptation and behaviour [31]. A more comprehensive approach that includes all sectors and collaborators is necessary. Community leaders on the other hand, play a critical role in shaping attitudes towards preventive measures. In areas where malaria education programs are effectively implemented, individuals are more likely to receive guidance on appropriate protective actions, such as using insecticide-treated bed nets, wearing protective clothing, and avoiding mosquito-infested areas (7, 15). If preventive measures are widely accepted and encouraged by community leaders, community members are more likely to comply [7, 32].

One significant discovery from the NGT session emphasizes ecotourism and activities involving monkeys—such as pet ownership or their use as food—as emergent yet underexamined factors influencing P. knowlesi malaria transmission. These activities enhance human interaction with natural macaque hosts and mosquito vectors that transmit the parasite, consequently increasing the risk of zoonotic spillover. Ecotourism in Southeast Asian forests, where P. knowlesi is endemic, attracts visitors to habitats abundant in macaques and vector mosquitoes, frequently lacking sufficient protective measures. Research indicates that travelers returning from forested regions in Thailand and Malaysia have contracted P. knowlesi, suggesting that evolving tourism patterns are a contributing factor [33, 34]. Hobbies involving close interaction with macaques, such as pet ownership or bushmeat consumption, increase the likelihood of contact with infected primates and their habitats. Interactions frequently take place in rural or recently deforested regions, where the conventional demarcations between wildlife and human habitats become indistinct. Research utilizing satellite data on land use in Malaysian Borneo has established a connection between the risk of P. knowlesi and villages situated near forest edges and oil palm plantations, indicating that both environmental and behavioral factors play a role in transmission [35]. Despite their importance, these human-driven activities are infrequently considered in malaria control strategies and literature, highlighting the necessity for additional research into their influence on zoonotic malaria dynamics.

Implications for further research

In Malaysia, where P. knowlesi has become the predominant malaria species, research should concentrate on understanding the socio-cultural and behavioural factors that affect malaria transmission and prevention behaviour. Occupational exposure, which usually focus among plantation workers, farmers, and loggers, is well-documented in most literatures. Nevertheless, new factors such as the impact of ecotourism and leisure activities may necessitate additional exploration [36]. In contrast to occupational risk, these activities may engage persons with limited knowledge on malaria and likelihood of adopting preventative measures is less. Future study may also seek to measure the degree to which ecotourism and recreational activities influence malaria risk and evaluate the efficacy of focused health education initiatives in mitigating exposure among non-occupationally exposed populations. Studies should not only evaluate exposure hazards but also investigate the cultural views and misconceptions regarding malaria within impacted groups [37]. Research through qualitative studies in examining traditional beliefs, disinformation, and socio-cultural obstacles to the adoption of preventive measures might yield significant insights for the development of culturally sensitive interventions [12]. Comprehending local communities' perceptions of malaria risk, treatment alternatives, and vector control methods might assist public health workers in formulating more effective communication tactics [36]. By tackling community-specific issues and utilizing traditional knowledge, initiatives can achieve greater acceptability and sustainability. Community involvement is an essential component of malaria management. Research should evaluate the influence of village leaders, religious authorities, and community groups in advancing malaria awareness and preventive initiatives [7]. Studies in Sabah, Malaysia demonstrated that involvement of village leaders and community health volunteers improves awareness and uptake of preventive practices in knowlesi-endemic areas [15, 37]. However, evidence on the specific influence of leadership endorsement in P. knowlesi prevention remains limited.

Research can also be directed to the exploration of environmental and ecological determinants of P. knowlesi transmission, especially in endemic or high-risk hotspot. There has been modelling done on the impact that deforestation, urbanization, and climate variability on the risk of malaria [3]. Nevertheless, there remains a gap in localized studies that examine how specific environmental factors shape disease risk and community behaviour in different regions of Malaysia and SEA. For example, there is limited literature examining how community adaptation to fragmented landscapes influences changes in macaque population dynamics [38]. Habitat alteration of macaque hosts and mosquito vectors secondary to anthropogenic activities brings them into closer contact with human populations [39]. These interconnected human–animal–environment interactions underscore the relevance of a One Health perspective in understanding P. knowlesi transmission [40]. The aftereffects of global growth, combined with geographical disadvantages and climate variability, are always associated with increased transmission risks. These environmental changes not only increase malaria incidence but also influence community responses, as rural communities reliant on farming and forest-related activities are more vulnerable. Socioeconomic determinants, including healthcare access and information inequity, further influence community behaviour towards preventive measures [41]. Disease interventions and public health strategies therefore need to incorporate these environmental, behavioural, and ecological determinants within a One Health framework to develop comprehensive, sustainable, and meaningful malaria control and prevention programmes, especially in resource-limited settings.

Multimodal approaches are required to manage zoonotic malaria, and prevention strategies cover a range of options. To manage vector-borne diseases (VBDs), the World Health Organization recommends a multi-sectoral strategy [42]. Understanding the sociocultural and behavioural aspects of local communities that interact with their societal and environment structure is necessary to design malaria programs that are both sustainable and effective. Sharing viewpoints and life experiences can motivate policymakers, stakeholders, and community members to make meaningful changes and act [43]. A useful theoretical, epistemological, and methodological framework for documenting and interpreting local issues of concern collaboratively and in-depth can provided by study such as community-based participation research (CBPR) [44]. Panel experts in this NGT study agreed that to deduce a relationship between human behaviour with health determinants, awareness and knowledge alone is inadequate. As a collection of techniques for recognizing and resolving regional issues of concern as well as a means of integrating social justice, equity, cultural humility, and mutual learning into research-community partnerships, CBPR has garnered significant traction especially in marginalised and rural communities [44]. Many malaria control programs have benefited from community-level activities that function as essential mechanisms enabling program implementation, even though community engagement has not historically formed a strong component of malaria control and elimination strategy [37, 45–47].

Strength and limitation

This NGT study provides a well-organized framework for expert consensus and structured approach to decision-making, thus promoting the sharing of knowledge and viewpoints. It ought to be regarded as a technique for collecting expert insights instead of delivering conclusive evidence. Nevertheless, this study identifies key P. knowlesi malaria risks and the behavioural factors that influence prevention and treatment practices in endemic communities, highlighting the need for a comprehensive research agenda that addresses these interconnected health determinants. This study also highlights several elements deemed necessary to be explored concerning malaria risk and its preventive behaviour among the community, underscoring the need of a comprehensive research agenda including a wide spectrum of determinant of health. This involves individual, interpersonal, organizational, community, policy, and societal level factors that can be explored more. In contrast to conventional survey techniques, NGT offers a dynamic environment where specialists can work together to enhance and rank essential topics via organized discussion and collective agreement. Considering lack of specific guidelines for P. knowlesi malaria prevention and control, this NGT study may presents a flexible strategy for pinpointing and prioritizing the key factors that affect prevention behaviour in the community. Utilizing the combined knowledge of experts experienced in malaria intervention and groundwork, this study may effectively address knowledge gaps in this evolving area. This approach also promotes the inclusion of local viewpoints especially among the health implementers and fieldworkers, guaranteeing that the factors identified, and the recommendations made are relevant and feasible for the communities impacted.

Despite all these positive insights, limitation of this NGT method is the potential for selection bias. The reliability and applicability of findings depend heavily on the composition of the expert panels. If the panel lacks diversity in expertise or fails to include representatives from all relevant sectors, certain perspectives may be underrepresented and may overlook critical determinants that influence malaria transmission and prevention behaviour. We tried to mitigate the issue by recruiting panels with various background of expertise and officers related to malaria research, prevention and control. The combination of researchers and implementers helped to balance the individual subjective judgements towards a group consensus. The decision to conduct the NGT session online enabled participation from panellists based in different locations and improved overall accessibility; however, it also reduced opportunities for informal interaction and the use of non-verbal cues that may enrich qualitative exchanges. Even so, the structured format of NGT supported clear discussion and consensus building. Future researchers using online NGT may benefit from interactive platforms and strong facilitation techniques, while in-person sessions may be preferable when deeper qualitative engagement is required.

Conclusion

Key elements influencing P. knowlesi malaria risk and prevention practices within at-risk communities were successfully identified and prioritized in this study using NGT. These findings emphasize the complex nature of malaria transmission and stress the necessity for tailored, context-specific therapies. The substantial consensus attained in a single NGT session underscores the efficacy of this structured method in consolidating expert perspectives for public health decision-making. Integrating these findings will enable malaria control efforts to be more effectively customized to meet the distinct problems encountered by vulnerable communities in Malaysia. Although some factors, like environmental and human-related problems, were known, more research is needed to make it easier for these vulnerable communities to understand and give their opinions. For that, future research may look into account how complicated the factors listed in this study are. This study presents a significant foundation for future research, providing a methodical approach to improve malaria preventive techniques and augment community engagement. Incorporating these insights into public health policies and intervention programs may enhance the efficacy of P. knowlesi malaria control. To maintain the pace of malaria elimination activities in SEA, further research should build on this consensus and address these factors using all-encompassing, context-specific methodologies that transcend traditional approaches.

Author contributions

Each author contributed significantly in this work.

Funding

Universiti Kebangsaan Malaysia’s Medical Research Ethics Committee (JEP-2024-504).

Data availability

Data is provided in supplementary information files 1. S1 Nominal Group Technique Voting Result.

Declarations

Ethics approval and consent to participate

This study received approval from the Medical Research & Ethics Committee, Ministry of Health Malaysia (project code: NMRR ID-24-02117-GPM (IIR); approval date: 15 August 2024) and the Universiti Kebangsaan Malaysia Medical Research Ethics Committee (ethics reference no: JEP-2024-504; approval date: 5 August 2024).

Consent for publication

All participants provided consent for publication.

Competing interests

The authors declare no competing interests.